15:00:18 <vishalmanchanda> #startmeeting horizon
15:00:18 <opendevmeet> Meeting started Wed Jan  5 15:00:18 2022 UTC and is due to finish in 60 minutes.  The chair is vishalmanchanda. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:18 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:18 <opendevmeet> The meeting name has been set to 'horizon'
15:00:58 <vishalmanchanda> hi anyone around for horizon meeting?
15:02:18 <vishalmanchanda> rdopiera: tmazur e0ne amotoki : ping
15:02:34 <tmazur> o/
15:03:33 <vishalmanchanda> Let's wait for 2 more minutes if more people joins us.
15:04:10 <rdopiera> o/
15:04:16 <amotoki> o/
15:04:43 <vishalmanchanda> ok let's start the meeting.
15:04:49 <vishalmanchanda> Welcome back everyone and happy new year!
15:04:59 <vishalmanchanda> I hope everyone is safe and healthy...
15:06:08 <vishalmanchanda> agenda for today meeting  https://etherpad.opendev.org/p/horizon-release-priorities ( at line 32)
15:06:21 <vishalmanchanda> #topic Notices
15:06:35 <vishalmanchanda> few announcement from my side for this week
15:06:48 <vishalmanchanda> As usual Yoga Schedule https://releases.openstack.org/yoga/schedule.html
15:07:07 <vishalmanchanda> This week is Yoga-2 milestone week.
15:07:33 <vishalmanchanda> Next announcement is about adding review-priority label in horizon Gerrit dashboard
15:07:56 <vishalmanchanda> Patch for the same is already merged.
15:08:34 <vishalmanchanda> So Core-reviewer please take a look at https://review.opendev.org/c/openstack/project-config/+/821934/1/gerrit/acls/openstack/horizon.config
15:08:34 <vishalmanchanda> to understand different value of review-priority label for those who don't know.
15:09:43 <vishalmanchanda> that's all announcement from my side for this week.
15:09:51 <vishalmanchanda> moving to next topic
15:10:08 <vishalmanchanda> #topic release-priorities
15:10:37 <vishalmanchanda> No update from my side for this topic.
15:11:18 <vishalmanchanda> If anyone would like to add anything about this topic, please go ahead.
15:11:42 <vishalmanchanda> Today I am investigating horizon-integration job failure.
15:12:08 <vishalmanchanda> You may have noticed few horizon-integration test are failing on your patches even after rechecks.
15:12:34 <rdopiera> I've seen that this has started with a timeout first
15:12:37 <vishalmanchanda> but If I look at horizon-integration periodic job it is passed today https://zuul.openstack.org/builds?job_name=horizon-integration-tests&project=openstack%2Fhorizon&branch=master&pipeline=periodic
15:12:40 <rdopiera> the failures came later
15:13:08 <vishalmanchanda> So I am confused that job is actually failing or not or just random test failure which usually happens in this job.
15:13:18 <rdopiera> perhaps the timed out job wasn't cleaned up properly
15:13:59 <tmazur> it looks like this
15:14:21 <tmazur> pagination tests usually fail if there wasn't a proper clean up
15:15:04 <vishalmanchanda> Could we do anything to fix it?
15:16:48 <vishalmanchanda> anyway I will push a patch on gate to investigate these pagination test failures
15:17:23 <vishalmanchanda> because If I run these integration test in my local env. it takes too much time.
15:18:06 <vishalmanchanda> If anyone know the command to run a single integration test please drop here that would be great help for me.
15:19:50 <amotoki> vishalmanchanda: perhaps the usual way to specify tests to be run can be used. {posargs} is passed to the test suite. Doesn't it work?
15:20:32 <amotoki> it seems it should work as long as I checked the env definition in tox.ini.
15:21:35 <tmazur> INTEGRATION_TESTS=1 pytest ./openstack_dashboard/test/integration_tests --ds=openstack_dashboard.test.settings -x -v -k test_images_pagination
15:21:45 <tmazur> that's what I'm using
15:22:01 <vishalmanchanda> amotoki: ohh yeah it's working
15:22:15 <vishalmanchanda> tmazur: thanks I will try that as well.
15:22:20 <amotoki> we have an example in our docs https://docs.openstack.org/horizon/latest/contributor/testing.html#py37
15:22:38 <amotoki> it is for py37 but pytest is used in the integration tests, so it works
15:22:54 <rdopiera> by the way, that SELENIUM_HEADLESS=False is actually interpreted as True, because "False" is a non-empty string
15:23:40 <rdopiera> amotoki: that example won't work since we use run_tests.sh that ignores the parameters
15:24:06 <rdopiera> it used to work when we used pytest directly
15:24:48 <amotoki> rdopiera: really? we use run_tests.sh before switching to pytest and I confirmed it worked after switching to pytest.
15:25:11 <amotoki> rdopiera: perhaps I missed something though
15:25:12 <vishalmanchanda> yes let's say if I want to run 'test_create_delete_group' and I use "tox -e integration openstack_dashboard/test/integration_tests/tests/test_groups.py::TestGroup::test_create_delete_group" command to run it.
15:25:39 <vishalmanchanda> It will run all integration tests in my local env.
15:25:52 <vishalmanchanda> just checked in my env.
15:26:13 <rdopiera> the command that tmazur gave will work
15:26:15 <amotoki> correction: we don't use run_tests. we use tools/unit_tests.sh.
15:26:55 <rdopiera> of course you have to run it in tox's virtualenv
15:29:34 <vishalmanchanda> ok moving to next topic
15:29:48 <vishalmanchanda> #topic Bug deputy report
15:30:18 <vishalmanchanda> We have 4 new bug from last 2 weeks.
15:30:35 <vishalmanchanda> Please take a look at all.
15:32:49 <amotoki> we repeatedly receive a bug report like https://bugs.launchpad.net/horizon/+bug/1956437
15:32:51 <vishalmanchanda> First one is look like a duplicate of some older bug but not sure exactly which one
15:33:08 <amotoki> the horizon document seems wrong
15:33:16 <amotoki> but I have no time to check the installation guide.
15:33:36 <vishalmanchanda> I can check and update it for ubuntu atleast.
15:33:41 <amotoki> do we want just to revert the configuration without checking what is actually required?
15:33:58 <amotoki> we have a report for RHEL/CentOS installation guide too.
15:34:04 <vishalmanchanda> will try to test and push a patch for the same asap.
15:34:07 <amotoki> s/a report/a similar report/
15:35:20 <rdopiera> looks like the guide linked shoud be fixed to say OPENSTACK_HOST = "controller:5000" instead of OPENSTACK_HOST = "controller"
15:35:22 <vishalmanchanda> rdopiera: tmazur : if you guys can check for the RHEL/CentOS that would be great
15:37:25 <rdopiera> same for rhel guide, although I doubt anybody uses that
15:38:33 <vishalmanchanda> For 2nd bug https://bugs.launchpad.net/horizon/+bug/1955773, a patch is already up.
15:38:41 <vishalmanchanda> #link https://review.opendev.org/c/openstack/horizon/+/822212
15:39:32 <vishalmanchanda> but I am not completely sure if above patch going to fix the issue
15:39:34 <amotoki> I am looking at it. I would like ot understand what really happens before giving +2
15:40:09 <vishalmanchanda> amotoki: great, thanks
15:40:21 <amotoki> I can reproduce it in my local env and I see a same object is referred repeatedly
15:42:11 <vishalmanchanda> I don't know I have pushed a depend-on patch in openstack/senlin-dashboard using horizon patch but doc job still fails.
15:42:28 <vishalmanchanda> See https://review.opendev.org/c/openstack/senlin-dashboard/+/822777
15:42:54 <amotoki> Depends-On does not work as horizon is isntalled from PyPI
15:43:11 <vishalmanchanda> amotoki: ohh I see
15:44:19 <vishalmanchanda> third bug https://bugs.launchpad.net/horizon/+bug/1955674
15:46:03 <amotoki> is it about Ussuri, right?
15:46:38 <vishalmanchanda> yeah if look at the bug summary.
15:46:46 <amotoki> If so, the latest nova policy with system-scope in nova does not work with horizon as horizon does not support it as of Ussuri.
15:47:23 <amotoki> more, deprecated policies are not supported either.
15:47:44 <rdopiera> and the reader role was not supposed to work in Ussuri either
15:47:46 <amotoki> IIRC the latter is the main reason we did not sync nova policy at that moment
15:48:25 <rdopiera> I think they are trying to use a feature that was not complete in that version
15:48:40 <amotoki> rdopiera: +1
15:48:43 <vishalmanchanda> Please add the comments in bug summary and close the bug.
15:48:49 <vishalmanchanda> amotoki: rdopiera : thanks
15:49:27 <vishalmanchanda> For 4th bug https://bugs.launchpad.net/horizon/+bug/1954619
15:49:54 <vishalmanchanda> I think amotoki  already added the comments and we can close this bug
15:51:34 <vishalmanchanda> that's all from my side for this week.
15:51:48 <vishalmanchanda> no topic in On-Demand Agenda
15:52:02 <rdopiera> I have a request for a topic
15:52:16 <vishalmanchanda> rdopiera: Please go ahead.
15:53:26 <rdopiera> I'm trying to fix a bug where a port creation/update fails with custom policies that don't allow the user to change the port security checkbox -- because we send the checkbox value even if it didn't change.
15:53:59 <rdopiera> I'm fixing it by adding a policy check for that particular form in the field, to hide the field when the user doesn't have permissions to use it.
15:54:17 <rdopiera> The patch I have so far is here: https://review.opendev.org/c/openstack/horizon/+/810224/
15:54:58 <rdopiera> However, from manual testing done by vishalmanchanda, it doesn't work correctly. I'm at a loss at how this policy check should be done properly, to match how it is done on the neutron side.
15:55:14 <rdopiera> I would like to request help with this, if possible.
15:59:58 <vishalmanchanda> rdopiera: I will test it again tomorrow and will check if I can help you.
16:00:11 <vishalmanchanda> ok we are out of time.
16:00:37 <vishalmanchanda> Thanks everyone for joining, See you next week!
16:00:56 <vishalmanchanda> #endmeeting